Elektronik Tidningen: The fiery soul who takes ideas to results

Broadband for everyone has been a strong driving force for Mats Andersson for more than 20 years.

Click for Pdf in Swedish / Article in English below

by Anna Wennberg, Editor, Elektronik Tidninger

– I have had that vision ever since I worked on Bill Gates’ Teledesic project.

The ambition in Teledesic was to get out broadband to the four billion people who did not yet have it. In the middle of the 1990s, broadband was just a few megabits. Today it is clearly more.

– This means that there are still four billion that do not have broadband and that the gap between poor and rich only increases because the internet is one of the most effective ways to create prosperity, states Mats Andersson.

Teledesic can be seen as a precursor to current satellite systems Starlink, Kuiper and Oneweb. Behind the company – planned to build over 800 low-flying satellites in similar tracks like Elon Musk (Starlink) and Jeff Bezos (Kuiper) talk about today – were the mentioned Bill Gates and also Boeing, the entrepreneur Craig McCaw, and a Saudi prince.

– There were about 30 companies that wanted to build the terminals for this project, and I led a team at Ericsson who was chosen to do it together with NEC.

Expensive terminals overthrew Teledesic

Right from the start, Mats saw the problem. To be able to follow satellites going from horizon to the horizon in  15 to 20 minutes the terminal antennas must be mechanical or electrically steered – and then the terminals become too expensive.

– Nobody believed it. They thought it would be possible to come down to $500 per terminal in large volumes. I had to explain that Ericsson already built point-to-point links which were much simpler, and that they were sold for $3,000 to $4,000 each, in the same volumes.

– It turned out that this is exactly what made the project shut down a few years later.

Mats still has the same fears that this applies to the terminals of the mentioned LEO systems, such as Starlink, Kuiper and Oneweb. These are satellites that are moving in relation to the earth at about 1,000 kilometers distance instead of the geostationary satellites (GEO) that are positioned at around an altitude of 36,000 kilometers.

– Many people think that the problem is to get the satellites up and build them cheaply. I think that the challenge is to build cheap enough terminals for the four billion that do not have broadband, he says and adds:

– Admittedly, I read the other day that one can buy Starlink terminals for $500, but I wonder how heavily they are subsidized. Besides, it still is a considerable amount of money for someone in the countryside in Africa or Asia.

At the same time, he is fully convinced that satellite communications today face the same expansion opportunity that mobile telephony had 30 years ago.

– When I changed jobs at Ericsson in 1990, I started at the “Department of Mobile  Telephony” in Stockholm. There was a department in Kista called that. Then there was talk about that cell phones might amount to something over time, but that the big thing for Ericsson was telephone exchanges and the fixed copper networks.

Since then, mobile communications have experienced exponential growth.

– I think it is about to level out. There is after all a mathematical certainty in that exponential curves flatten out. Independently, it is too expensive to expand mobile networks to the entire population of the Earth – if it was possible, it would have been done, reasons Mats.

A satellite has a completely different coverage area. Hybrid technology can be the solution. Seven years ago, Mats approached the solution and a new challenge – as he stepped into Almi Invest and asked where he can be most useful. He had just become financially  independent after a successful run during his time at the Gothenburg company Bluetest, which tests the antenna and radio performance.

– Almi suggested Forsway, so I got in there as Chairman of the Board in 2014. The company had the solution to what I discovered was the problem when I worked with Teledesic – to make cheap satellite terminals.

Today, the Skövde company Forsway makes the world’s most cost-effective satellite terminals for GEO satellites. The secret is hybrid technology. That means that the downlink goes via satellite, while the uplink uses a simple modem via 2G, 3G or ADSL.

Last year, Forsway did field tests in Arizona. There are 5-10 million people in the United States –  ADSL users who only have one Mbit uplink and down link. With Forsway’s system, they easily got 25 Mbit in the downlink, while ADSL was used for the return link.

The testers immediately accessed both video and got good connections to web pages. Right now, Forsway have commercial rollouts of their modems in the Philippines, Indonesia, and several countries in Africa. The company has also applied for a patent on hybrid technology for LEO systems. So far, it has been approved in Israel, while it is being processed in Europe and the United States.

– In the long run, I think that hybrid technology can pave the way for GEO terminals under $50 and LEO terminals under $200. We will see if it is possible to develop a good relationship with any of the companies that make LEO systems. Currently, they appear to work entirely internally with their own terminals, which feels ancient.

– Instead, they should create a standard, exactly as with the cell phone industry. Only then will it be  possible to bring down the cost of the terminals.

Bluetest is the most fun memory

If the vision to connect everyone who does not have internet is what Mats Andersson is focused on  right now, Bluetest is still the most fun memory of his versatile career.

– Yes, I would say that. Even if Ericsson has been a fantastic company to work in, and good  schooling before starting at a start-up company like Bluetest with only two employees.

The Bluetest era began in 2006 when Ericsson wanted to rejuvenate. Then Mats Andersson had worked there for 20 years with everything from designing the antennas for a French space shuttle, developed Ericsson’s first HLR (Home Location Register) which kept track of all mobiles and services, sold-in a Swedish-developed antenna elements to Hughes Aircraft that was a leading American company in space and defense industry, to developing antenna elements for the Global Star satellite system.

He had also had time to roll out broadband in rural areas in Sweden and other parts of Europe. And just before Ericsson offered 18 months’ salary to everyone who quit voluntarily, he was again at the antenna department at Ericsson Research to start looking at multiple antennas (MIMO) for 4G base stations. It was still 5-6 years before 4G broke through.

Takes ideas to results

INTERVIEW: Half of the world’s population does not have broadband of 10 Mbits or more. Mats Andersson wants to change that – and he is convinced that the solution is satellite communication where the terminal on the ground is the big challenge. At the same time, he offers a fascinating work life, which oscillates between satellite and mobile communications, managerial and engineering jobs as well as working with large and small companies.

“I wanted to do something similar and get at least ten times more capacity in the mobile network and at the same time reduce the output power ”

– But I took Ericsson’s offer. I had just turned 50 and I thought that it could be fun to try something new. I thought of starting by renovating the house, he laughs.

– I did not even have time to leave Ericsson before Professor Per-Simon Kildal at Chalmers called.

He had heard the rumor and needed to a CEO for Bluetest that he had operated for six years with a million in loss per year. The job interview was to attend an antenna conference in New Mexico, USA, to evaluate the interest in Bluetest’s product with potential customers.

– People came from Motorola and Nokia. I realized that there was an opportunity to sell it, so I took the job.

Took speed with NTT Docomo

The start was slow. The Bluetest chamber does not measure antenna diagrams, but measures antenna efficiency extremely fast. Admittedly, antenna diagrams are not that important for small antennas that are usually omnidirectional, but there was a conservative view among antenna engineers that antenna diagrams are needed.

– It was difficult to convince potential customers that the product works. But then I met a curious person from NTT Docomo, who is Japan’s largest mobile operator.

The meeting led to NTT Docomo sending a number of mobiles to Bluetest without revealing the result it got in its echo-free chambers where it measured antenna diagrams to calculate antenna efficiency. Bluetest measured and sent back the result. New phones came, and after a year back and forth came the first order.

– Then we had a breakthrough in Japan and after another year, NTT Docomo told providers of mobile phones that they should use our test chambers. Then it picked up speed.

At the same time, Mats realized that the company’s test chamber had repeatable reflexes with incredibly high accuracy, which opened for evaluating products with multiple antennas.

By placing antennas in different positions in the chamber, and then test, it was possible to optimize the location of the antennas.

– Previously, one had to measure antenna by antenna in echo-free chambers. It could take one hour per antenna. Then one calculated what it would be, and it did not turn out very well. It took at least half a day to estimate MIMO capacity for a router. We did it in one minute.

Bluetest became first in the world to be able to do repeatable 4G measurements. This led to everyone who developed 4G mobiles needing a Bluetest chamber.

Today Bluetest has 60 employees and Mats is still a co-owner. However, he left the CEO position ten years ago when growth began to go steadily on the way up.

– I saw an ad in Göteborgsposten that Huawei was looking for a manager for its Gothenburg office, so I applied.

The Chinese liked that Mats was 55 years old and had long experience. He got the job and a fair amount of freedom –  the task was to deliver ideas on how the next generation base station should look like.

– Then I thought back to the satellite antenna which I worked with in the 90s, with 127 antenna elements that gave 30-40 times more capacity than a single antenna. I wanted to do something similar and get at least ten times more capacity in the mobile network and at the same time reduce the output power.

At that time, a base station had at most four antennas. Mats suggested 48.

Gained trust with Huawei

The idea was not  rceived well. The Chinese were skeptical and thought it was too complex and risky. But Björn Sihlbom at  Huawei in Gothenburg thought the idea was interesting. In the end, Mats got a small budget and a year together with a small team in Gothenburg to show that the idea had merit.

– When we started the project, we had not heard about Massive MIMO, which we  later realized it was  called in academic articles.

The development in Gothenburg led to Huawei becoming the first in the world to develop base stations with Massive MIMO. The first was delivered to China Mobile in 2015, the year after that Mats left the company.

– Per-Simon Kildal gained significant dividends from Bluetest in 2014, and I was rewarded as well. Then I realized I did not need one full time job anymore, so I quit Huawei.

Instead, Mats stepped into Almi Invest in 2014, which five years earlier actually made its very first investment in just Bluetest. There he offered his competence as a mentor, and nowadays he is, among other things, involved in the start-up companies Forsway, Satcube and Brinja.

The Gothenburg company Satcube makes the world’s most portable satellite terminals, while Brinja develops IoT systems for construction sites.

– Now I want to devote myself to my start-up companies. I also study Russian literature part-time at the university, and I am thinking of starting to learn Russian in the autumn as a  hobby, he says, pleased.

He wanted to understand the universe

Mats Andersson’s interest in space began at the age of five, in the middle of the space race between then The Soviet Union and the United States. A space race that started seriously already four years earlier – on October 4, 1957 – the Soviet satellite Sputnik was launched into orbit around the Earth.

– I wanted to be an astronomer, my mom told me a lot about the sky. Most people thought that I wanted to be an astronaut. Everyone talked about it, but I wanted to study the stars, not to go up in space.

The interest in space continued throughout school. But when it was time to choose university studies, the choice fell on a Master of Science in Engineering Physics at Chalmers, as the labor  market for astronomers was not very optimal.

– During my last year at Chalmers, I probably had half of my subjects in astronomy or quantum physics, as I had not completely given up the idea that become an astronomer.

The next step in his career was Onsala Space Observatory, there he received a doctoral position.

– I could not start immediately. Instead, the professor who would be my supervisor arranged a job for me at Ericsson’s antenna department in Mölndal. There I got to be a part of designing the reflector antennas that would sit on the Tele-X satellite. Tele-X was the first TV and datacom satellite for  the Nordic countries.

Six months later Mats – who was still completely set on becoming an astronomer – received a doctoral position.  2.5 years later he earned his licentiate degree.

– At that time I had started to get a little disillusioned with research. It was lonely and I liked working in large teams, as at Ericsson.

– So, in 1986 I decided that an academic career was probably not for me. I called Ericsson and asked for a job, which I got.

ANNA WENNBERG